From: Tom London <tbl@research.uucp>
Date: Mon, 6 Nov 1978 10:22:40 +0000 (-0500)
Subject: Bell 32V development
X-Git-Tag: Bell-32V~684
X-Git-Url: https://git.subgeniuskitty.com/unix-history/.git/commitdiff_plain/54008ed01ad000996d314b3e52611af31c2647a1

Bell 32V development
Work on file usr/src/libm/erf.c

Co-Authored-By: John Reiser <jfr@research.uucp>
Synthesized-from: 32v
---

diff --git a/usr/src/libm/erf.c b/usr/src/libm/erf.c
new file mode 100644
index 0000000000..b724b905b4
--- /dev/null
+++ b/usr/src/libm/erf.c
@@ -0,0 +1,116 @@
+/*
+	C program for floating point error function
+
+	erf(x) returns the error function of its argument
+	erfc(x) returns 1.0-erf(x)
+
+	erf(x) is defined by
+	${2 over sqrt(pi)} int from 0 to x e sup {-t sup 2} dt$
+
+	the entry for erfc is provided because of the
+	extreme loss of relative accuracy if erf(x) is
+	called for large x and the result subtracted
+	from 1. (e.g. for x= 10, 12 places are lost).
+
+	There are no error returns.
+
+	Calls exp.
+
+	Coefficients for large x are #5667 from Hart & Cheney (18.72D).
+*/
+
+#define M 7
+#define N 9
+int errno;
+static double torp = 1.1283791670955125738961589031;
+static double p1[] = {
+	0.804373630960840172832162e5,
+	0.740407142710151470082064e4,
+	0.301782788536507577809226e4,
+	0.380140318123903008244444e2,
+	0.143383842191748205576712e2,
+	-.288805137207594084924010e0,
+	0.007547728033418631287834e0,
+};
+static double q1[]  = {
+	0.804373630960840172826266e5,
+	0.342165257924628539769006e5,
+	0.637960017324428279487120e4,
+	0.658070155459240506326937e3,
+	0.380190713951939403753468e2,
+	0.100000000000000000000000e1,
+	0.0,
+};
+static double p2[]  = {
+	0.18263348842295112592168999e4,
+	0.28980293292167655611275846e4,
+	0.2320439590251635247384768711e4,
+	0.1143262070703886173606073338e4,
+	0.3685196154710010637133875746e3,
+	0.7708161730368428609781633646e2,
+	0.9675807882987265400604202961e1,
+	0.5641877825507397413087057563e0,
+	0.0,
+};
+static double q2[]  = {
+	0.18263348842295112595576438e4,
+	0.495882756472114071495438422e4,
+	0.60895424232724435504633068e4,
+	0.4429612803883682726711528526e4,
+	0.2094384367789539593790281779e4,
+	0.6617361207107653469211984771e3,
+	0.1371255960500622202878443578e3,
+	0.1714980943627607849376131193e2,
+	1.0,
+};
+
+double
+erf(arg) double arg;{
+	double erfc();
+	int sign;
+	double argsq;
+	double d, n;
+	int i;
+
+	errno = 0;
+	sign = 1;
+	if(arg < 0.){
+		arg = -arg;
+		sign = -1;
+	}
+	if(arg < 0.5){
+		argsq = arg*arg;
+		for(n=0,d=0,i=M-1; i>=0; i--){
+			n = n*argsq + p1[i];
+			d = d*argsq + q1[i];
+		}
+		return(sign*torp*arg*n/d);
+	}
+	if(arg >= 10.)
+		return(sign*1.);
+	return(sign*(1. - erfc(arg)));
+}
+
+double
+erfc(arg) double arg;{
+	double erf();
+	double exp();
+	double n, d;
+	int i;
+
+	errno = 0;
+	if(arg < 0.)
+		return(2. - erfc(-arg));
+/*
+	if(arg < 0.5)
+		return(1. - erf(arg));
+*/
+	if(arg >= 10.)
+		return(0.);
+
+	for(n=0,d=0,i=N-1; i>=0; i--){
+		n = n*arg + p2[i];
+		d = d*arg + q2[i];
+	}
+	return(exp(-arg*arg)*n/d);
+}